Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Nathan Riddle]

So I have the privilege of hanging out with a friend on Sunday and helping come up with some limiter/EQ settings for his SM80/TH115 system. We’re also going to compare other speakers and generally ‘geek out’ have a fun day.

PSW members are welcome to join. PM me lower GA

The purpose of the day is two fold; system tuning & system comparison.
-My primary goal is to get him up and running 100%
-After that we’d then have some fun listening to the different speakers.

I’ve been reading the ARTA/REW manuals & refreshing myself on the techniques. (I know, I know... I need to take a formal SMAART class, and Pat’s OptEQ class. $$$).

I think I understand everything well enough to do a good job, not as good as you all, but at least a good solid job.

I’m looking for some tips & tricks gained only through experience that might help me along the way. For instance:
I remember some of what Mark said in my TD1 thread about setting the HPF extremely steep so they all have similar responses.

My general idea relating to system setup is:
Setup everything (SM80, TD1, JTR, 835p)
A) measure/test everything so it is in good working order
B) apply proper limiting that balances performance vs safety & test/verify that limiting
C) set generic xover & delay for an ideal scenario
D) apply corrective EQ as needed so the system is linear/flat

My general thought in regards to listening to different speakers:
1) measure each speaker so they have the same low cutoff freq
2) set volume to same
3) listen?
OR
1) measure each speaker so they have the same low cutoff freq
2) EQ so they all have the same trace
3) set volume to same
4) listen

Any tips/thoughts on all this?

[Luke Geis]

I think you may be over complicating it a little. I prefer the listen first, asses, then correct ideology. Also, limiting and system protection is the last thing you do, not the first. Think of it like this. System design, integration, and optimization.

You design the system first. This is done by selecting the speakers, amps, DSP and deployment technique. This is followed by actually putting that plan into action. Lastly, you do the corrective measures and end with system protection.

Once you get to the last step of optimization, you have to have assessed what the problems are and what measures you can do to fix them and if you can even fix them. Remember things like comb filtering from room modes, or bad time alignment are not correctable with EQ. So you have to be able to listen and determine what the problem is. The Smaart type systems help a bit with sussing out some of the obvious issues. The end all be all though is listening and bias confirmation. If you hear it and Smaart confirms it, then it does exist and you can correct. If you hear it and Smaart doesn't confirm it, then you have to find another method of testing or simply rely on your ear or Smaart as the answer. Sometimes moving a speaker 2-3' can make a difference in a positive way and sometimes there is nothing you can do and you have to live with what you got.

Measuring speakers is probably the least concerning thing to me. There is probably always a little bit of hysteresis from one speaker to the next ( of the same model )and that little bit of difference isn't necessarily a bad thing. If each speaker is a little bit different, you can reduce comb filtering and destructive interference a little actually. When I measure a system I am looking for two predominant things. Time alignment and linearity. Now linearity is sort of a misnomer really. The system is likely to never be 100% board flat from 20hz -20khz. And I would not expect there to be as much energy at 16khz as at 160hz. What I do not want to see is no large humps or dips in the frequency response. Now if 1 or 2 do exist, the question is what is causing it and is it an expected outcome? If there is a large dip at 8khz and the cohesion at 8khz is low in Smaart, I will negate that result as being a false reading if my ears are telling me that 8khz is there just fine. Again, listen first. Now if there is a large hump around 160hz and I expect it to be there ( due to experience or room and system setup that supports that possibility ) I will use corrective EQ to dial it out. All said and done I am just looking to see a smooth frequency response that matches what I', looking for and what I expect to see. I am not as worried as I mentioned if L & R are perfectly EQ'd the same. I guess my sequence would be this:

1. Listen to the DUT ( the speaker )

2. Asses any problems and probable causes.

3. With measurement software, confirm/deny biases and assessments.

4. Make corrective adjustments with either physical placement or DSP corrections to improve results and repeat steps 3-4 until what you hear is what you see, or you can definitively rely on your ear or the measurement software.

5. Set system protections to manage the potential oopsies. Honestly, with turnkey systems, this is usually less of an issue. If you use the proprietary amps and or recommended DSP that is associated with your speakers, this step is generally already done for you. More so with self-powered units and top-shelf units that have such options. If you are running off brand amps with X speaker, refer to the vendors recommended settings first, and if not available then go into deep editing mode and find the setting that works for you.

Some things to consider. Speakers tend to have a phase wrap at or near their lower tuning frequency where the ports start to work. This is likely why many vendors set the crossover points nearer 100-120hz. An attempt to reduce the lack of low-end control in that speaker and help keep some headroom. I prefer running my mains as low as practical because I also prefer running low crossover points on the subs as well. I typically use a crossover point between 60-80hz for the subs depending on what's going on and the needs. This usually puts the mains around 80-100hz to acoustically match with the subs. How steep of a filter is more based on listening and deciding what sounds best. Steep is great but can get real thin sounding really quick with higher crossover settings between 100-120hz. Some speakers do well with really steep filters and others do not. What sounds right? What gives me the smoothest most linear outcome and is within the speaker's optimal performance regions.

[Nathan Riddle]

Hey Luke,

Thanks for the response! Very informative and helpful

Let’s get to it.

I think you may be over complicating it a little. I prefer the listen first, asses, then correct ideology. Also, limiting and system protection is the last thing you do, not the first. Think of it like this. System design, integration, and optimization.

Perhaps a little, but I’ve read too much about how psychoacoustics & poor technique influences casual speaker comparison meet ups. I want to eliminate that as much as possible (or at least know what they are and attempt to remove them).

I disagree that limiting is the last step. It is the first, I want to protect what I am about to ‘play’ with. We are talking protection of the drivers, not compression of the system for taste.

You design the system first. This is done by selecting the speakers, amps, DSP and deployment technique. This is followed by actually putting that plan into action. Lastly, you do the corrective measures and end with system protection.

Technically, if the system is deployed with manufacture gear/presets then the corrective & system protection is done first, the last step is room (or air absorption) EQ.

Once you get to the last step of optimization, you have to have assessed what the problems are and what measures you can do to fix them and if you can even fix them. Remember things like comb filtering from room modes, or bad time alignment are not correctable with EQ.

So you have to be able to listen and determine what the problem is. The Smaart type systems help a bit with sussing out some of the obvious issues. The end all be all though is listening and bias confirmation. If you hear it and Smaart confirms it, then it does exist and you can correct. If you hear it and Smaart doesn't confirm it, then you have to find another method of testing or simply rely on your ear or Smaart as the answer. Sometimes moving a speaker 2-3' can make a difference in a positive way and sometimes there is nothing you can do and you have to live with what you got.

Agreed, a good refresher. I want to balance ear vs measurement and find a best compromise.

And I’m glad you reiterated all that for anyone following along at home. I am aware of the pitfalls associated with measurement systems, it’s only as smaart as the user (pun intended)

Measuring speakers is probably the least concerning thing to me. There is probably always a little bit of hysteresis from one speaker to the next ( of the same model )and that little bit of difference isn't necessarily a bad thing. If each speaker is a little bit different, you can reduce comb filtering and destructive interference a little actually. When I measure a system I am looking for two predominant things. Time alignment and linearity. Now linearity is sort of a misnomer really. The system is likely to never be 100% board flat from 20hz -20khz. And I would not expect there to be as much energy at 16khz as at 160hz. What I do not want to see is no large humps or dips in the frequency response. Now if 1 or 2 do exist, the question is what is causing it and is it an expected outcome? If there is a large dip at 8khz and the cohesion at 8khz is low in Smaart, I will negate that result as being a false reading if my ears are telling me that 8khz is there just fine. Again, listen first.

Now if there is a large hump around 160hz and I expect it to be there ( due to experience or room and system setup that supports that possibility ) I will use corrective EQ to dial it out. All said and done I am just looking to see a smooth frequency response that matches what I', looking for and what I expect to see. I am not as worried as I mentioned if L & R are perfectly EQ'd the same.

While this is all a nice refresher I’m not sure what all that has to do with measuring a speaker and applying direct field EQ so that when placed in a room it is already producing a relatively flat/linear response.

The system tuning idea is to provide a blank slate for when taken to a gig/venue can then be molded to fit.

OptEQ methodology:
Layer 1 is direct field optimization that produces a reference response for the loudspeaker that confirms that it is functioning as designed and to its fullest potential.
Layer 2 addresses anomalies that result from placement of loudspeakers near room boundaries and other loudspeakers. Good choices here can make a system much more equalize-able.
Layer 3 addresses room issues such as low frequency build-up and resonances, using a dual-domain (time/frequency) approach.

Measuring for the sake of measuring, gaining skills, learning more about the process, etc

Also, I’ll Be measuring in a field and in the air so I can remove strong reflections.

I guess my sequence would be this:

1. Listen to the DUT ( the speaker )

2. Asses any problems and probable causes.

3. With measurement software, confirm/deny biases and assessments.

4. Make corrective adjustments with either physical placement or DSP corrections to improve results and repeat steps 3-4 until what you hear is what you see, or you can definitively rely on your ear or the measurement software.

5. Set system protections to manage the potential oopsies. Honestly, with turnkey systems, this is usually less of an issue. If you use the proprietary amps and or recommended DSP that is associated with your speakers, this step is generally already done for you. More so with self-powered units and top-shelf units that have such options. If you are running off brand amps with X speaker, refer to the vendors recommended settings first, and if not available then go into deep editing mode and find the setting that works for you.

I agree except for the previously stated stance of performing protection first. I want to be able to run the system hard without fear during my ‘testing’.

Some things to consider. Speakers tend to have a phase wrap at or near their lower tuning frequency where the ports start to work. This is likely why many vendors set the crossover points nearer 100-120hz. An attempt to reduce the lack of low-end control in that speaker and help keep some headroom. I prefer running my mains as low as practical because I also prefer running low crossover points on the subs as well. I typically use a crossover point between 60-80hz for the subs depending on what's going on and the needs. This usually puts the mains around 80-100hz to acoustically match with the subs. How steep of a filter is more based on listening and deciding what sounds best. Steep is great but can get real thin sounding really quick with higher crossover settings between 100-120hz. Some speakers do well with really steep filters and others do not. What sounds right? What gives me the smoothest most linear outcome and is within the speaker's optimal performance regions.

————

To give some info about the setup.

Passive system, I have the manufacture presets/suggested settings.
Generic amps (PLD 4.5) so a decent amount of tuning available, but not the same as the DNA amps that Danley has.

Pretty sure I can take care of the system tuning myself. I’m not as confident in the finer details of xover & slopes, but I can brute force it.

I’d like more details about listening comparisons. I know don’t listen to tonality because that is EQ. But listen to other things, coverage, consistency, etc. But what about the ‘sound’ of the speaker (having taken into the tone into account via corrective EQ).

I’ll re-read the threads on SM80/THxxx systems for a starting point. Not worried about that stuff.

[Luke Geis]

The reason I say that system protection is done last is for two reasons.

1. It is what you would do after you have set crossovers and time alignment and since you are doing those two things, you may as well do your system EQ at the same time instead of going back to it.

2. System protection settings typically require that you run the system at performance levels in order to hear what those protections may be doing. There is no sense in killing your ears and the neighborhood before you go back to a lower level and then EQ it to a linear state and make it sound pretty.

Doing system protection settings first is like putting the cart in front of the horse. A typical setup looks like this assuming you are starting 100% from 0.

1. Design the PA.

2. Build the PA, this includes all connections from the mixer to the speakers.

3. Test the PA to be sure all works as it is intended L is L, R is R and subs are subs etc...

4. Optimize the PA by setting crossovers, time alignment, EQ and then you run the system up to performance level. While at performance level you can set protections and fine tune.

Your idea of doing the safety stuff first makes sense, but it just goes out of order. If you don't know where the end of the road is with a PA system, then you have to resort to common sense and the listen first, asses and then correct mantra. If you don't know how much is too much, then be cautious, look at the blinking lights and use your ears. Go slow and build up to performance volume. Hopefully, by the time you get there, the system will sound great and your ears won't be shot.

I glazed over the Linearization of a speaker bit. What I should have included was that if the speaker is linear or not before going into a new venue is of little importance in the long run. The idea that the speaker is flat and therefore you are just overlaying another EQ to correct for what the speaker does in the room, is great, but if you are doing what you should be ( which is using Smaart or other measurement software ) then you will already have the tools out to correct for that stuff anyway. You should be setting crossovers ( because the acoustic crossover points change with different setups ) and you will also need to time align, so you may as well do a system EQ then too. This then means that you can confirm that your speakers are linear in THAT SPACE. If you are not going to go through those steps, then I guess by all means have a " flat " setting and trust it. The big trust factor is that the setting you made was, in fact, good and that the speaker is as linear today as it was the last time you measured it. My big thing is that I don't like stacking EQ's and I don't like doing things twice. I prefer having the system sound the way it does without any EQ at all and then making all the EQ adjustments and corrections all at once. It just weeds out any what-ifs and other possible mistakes. Plus, if your speaker is set up next to a wall, that flat EQ setting is going out the window anyway.

It makes a little more sense for monitors because it does nail down the big variables, but in the end, you still have to go back and tune it, so you may as well do it from scratch.

[Mark Wilkinson]

Hi Nathan, looking forward to hearing how it went. 
Very aggressive agenda...I have a hard time comparing only 2 well lol.  Good luck !

My previous comment about using steep xover hpf's for mains, to make comparisons more valid, get's kinda impossible when all the mains can't reach down similarly. 
I'm guessing the SM-80 made it a little tough to pick a common low freq.....
I usually use 100Hz, LR8  out of pure habit.
 
FWIW,One cool thing about using linear phase crossovers is you can more easily slide x-over freq up or down and keep the sub to main timing/ phase alignment intact. 
This let's you use the same subs with different mains, to find the best x-over freq for each main, and still be aligned.
Allows much easier comparisons of mains with very different low end extensions..

[Mark Wilkinson]

It makes a little more sense for monitors because it does nail down the big variables, but in the end, you still have to go back and tune it, so you may as well do it from scratch.

Hi Luke, I can see where you're coming from, especially if you're a one man operator providing mixing as well.

About everything i do falls in to one of two categories:
either I provide speakers, processing, and amplification for audience coverage
or I tune providers speakers, processing, and amplification for audience coverage.

Either way, my goals are always the same:
flat magnitude
phase as flat as possible
coverage as even as possible

Then turn the wheel over to the BE...or provider if they're doing the mixing.
Levels between mains, fills, aux subs, is all theirs.....(but neither provider or BE or gets to touch processing settings if it's my gear!)
They are free to tune to room, tune to taste....knowing that they started with a flat mag and phase system..kinda the meyer approach I think...

When I also do the room tuning, or tune to taste, it's so nice to start with a known flat system...
let's me quickly figure out where problems are...and if it can be fixed...

So I guess I'm saying, uh-uh, I ain't going back to scratch unless i have too 

[ProSoundWeb Community]